Written evidence submitted by Dr Kristen MacAskill
C.i. Do not limit system boundaries to physical asset performance. To determine a ‘level’ of resilience we need to be clear on what constitutes a system and the outcomes we seek to influence; metrics or standards for the design of physical infrastructure do not necessarily ensure that infrastructure is governed well. Critical National Infrastructure (CNI) needs to be considered as a socio-technical system, acknowledging that infrastructure performance is influenced by decision-making capacity. Currently, organisational analysis of CNI is mainly limited to qualitative systems analysis that has limitations in replicability for comparative or oversight purposes. Experimentation is needed to develop these methods for more widespread adoption.
C.ii. There is currently no widely adopted process for assessing climate adaptation of infrastructure in design, but existing schemes and codes of practice could be revised to support adaptation efforts more effectively and transparently.
C.iii. Governance of CNI resilience can benefit from an organisation with a remit to ‘broker resilience’. The development of roles and responsibilities across different levels of government in Queensland, Australia, provides an example of the evolution of roles and responsibilities to respond to extreme climate events. The Queensland Reconstruction Authority (QRA) was set up as part of a response to a major state-wide emergency to fill a gap in governance capability, but over time its remit expanded to what can be described as a ‘resilience broker’ across different levels of governance. The QRA’s strength is that it connects the different levels: it provides assistance and distributes money to the local level, and it monitors how this is spent (reporting to national government).
C.iv. Technology development will undoubtedly aid CNI management, but human factors in governance and operations must receive similar attention. Any new technology must be considered within the existing context, that is, how the organisation makes decisions, how systems fit together, who has access to information, and who has mandated responsibility. Otherwise, it can fail to deliver results if the costs for wider change and upskilling are not accounted for.
1.1. This question is similar to that of the House of Lords Select Committee, who asked: Given the range of possible national risks, and the need to achieve a balance between efficiency and resilience, what level of assurance should the Government be seeking on the UK’s resilience to hazards? What would effective national risk management achieve, and how could its success be measured? In response to that inquiry, I emphasised that “Compared to standard risk management, a resilience paradigm promotes greater focus on developing interventions that might help a system respond and recover in more extreme scenarios, accepting that it might not be feasible to directly treat the risk”. The House of Lords Select Committee final report covers the issue of managing short- versus long-term risks well. I will underline here the importance of managing CNI resilience not just through design of the physical asset but in how it is governed.
1.2. To determine a ‘level’ of resilience we need to be clear on what constitutes a system and the outcomes we seek to influence; metrics or standards for the design of physical infrastructure do not necessarily ensure that infrastructure is governed well. There is debate over how to determine what a ‘level’ of resilience means, which is influenced by how we frame the problem. Many argue that quantitative targets are needed so that performance can be directly measured. Engineering design standards focus on the physical resilience of infrastructure, that is, the asset’s ability to withstand/recover from shocks and stresses. There are two important considerations here: (A) The acceptable level of resilience that is designed into infrastructure evolves over time based on prevailing professional wisdom and associated policies. As a result, there are legacy issues where existing infrastructure is not required to meet current design standards. (B) CNI needs to be considered as a socio-technical system, acknowledging that infrastructure performance is influenced by decision-making capacity. Resilience in this sense can be difficult to precisely quantify, for example, whether the capability exists to respond to events in a coordinated way. This leads to thinking about the organisation managing the assets.
1.3. Currently, organisational analysis of CNI is mainly limited to qualitative systems analysis that has limitations in replicability for comparative or oversight purposes. Researchers at the US Army Corps of Engineers have led some advances in the field,  but useful, consistent metrics for understanding a CNI system beyond its physical components remain elusive. Rasmussen and subsequently Woods (research specialists in system safety and human factors) highlighted how accidents emerge through “drift” or “migration” of organisational behaviour towards unsafe boundaries. This concept was used by Leveson in the development of the Systems-Theoretic Accident Modelling and Processes (STAMP) concept. STAMP involves a process that analyses organisations as interactions among people, societal and organisational structures, and engineering and physical assets. STAMP has specifically been applied to mapping control and oversight arrangements in complex systems; it has demonstrated how changes within these systems affect the originally designed oversight mechanisms, serving as a contributing factor to major accidents (such as water contamination in a water supply scheme).
1.4. There is a wider field of research and guidance on characteristics of a “resilient organisation” (e.g. British Standard 65000:2014). These tend to highlight generic features of an organisation which broadly apply to CNI, but do not provide advice specific to CNI or climate adaptation. It is also generally emphasised that organisational resilience is a relative concept only and cannot be reduced to a definitive indicator, only broadly categorised.
1.5. Experimentation is needed to develop these methods for more widespread adoption. The UK’s National Infrastructure Commission made recommendations in their 2020 study Anticipate, react, recover – resilient infrastructure systems, reflecting the need to set clear, transparent standards. Their report also refers to changes in wider governance that are required to support a more strategic shift (e.g., duties to support long-term investments, implementation of new stress testing processes, economic assessment that values resilience).
2.1. There is currently no widely adopted process for assessing climate adaptation of infrastructure in design, but existing schemes and codes of practice could be revised to support adaptation efforts more effectively. While the current parliamentary inquiry is not directly addressing building design, a recent review of sustainability assessment schemes (BREEAM, Green Star, and LEED) helps to demonstrate this point (similar assessment schemes exist that cover other infrastructure sectors). These schemes were developed in the private sector (the earliest emerging in the 1990s) to celebrate and advance best practice sustainable design. They are based on a graded award/certification structure: lower-level awards typically represent meeting a minimum level or code compliant efforts, the highest level reflects best practice and adoption of innovative techniques to respond to sustainability challenges. Climate change is among the issues addressed in the assessment.
2.2. The review highlighted that there has been a greater focus of research and policy on mitigation, compared with adaptation in infrastructure development. As such, the incorporation of climate adaptation into decision-making is less well advanced. While the sustainability assessment schemes are not mandated standards, we used them as a proxy for wider best practice in industry. In each scheme there was a clear emphasis on mitigation over adaptation assessment and often adaptation-related efforts are not included in mandatory criteria. This does not send a clear signal to investors – they cannot be certain to what extent climate risk has been addressed, even if a high “sustainability” rating has been achieved.
2.3. I have also supervised master’s research (unpublished) into defence and health sectors in the UK which has explored the systemic issues faced in considering longer-term implications of decisions. Those with responsibility to make decisions related to asset development do not have the remit to proactively incorporate or prioritise solutions to adapt to a changing climate. For example, a study of hospital estate management in England (completed in 2021) demonstrated that retrofitting to address climate change risk is predominantly a reactive practice that follows the experience of an extreme weather event, rather than part of an anticipatory, planned investment process. Funding exists predominantly to deal with backlog maintenance and urgent clinical needs.
3.1. The development of roles and responsibilities across different levels of government in Queensland, Australia, provides an example of an evolution of roles and responsibilities to respond to extreme climate events. I provide this as an example of the evolution of capacity over time to support resilience-building in the community. This is based on a study conducted for the Royal Academy of Engineering as part of its Safer Complex Systems programme (due for publication this year). We reviewed the evolution of disaster risk governance in Queensland, Australia, focusing on the role of the Queensland Reconstruction Authority (QRA). While there are context-specific circumstances that led to the creation of the QRA, the case offers general learning for governance of disaster risk and building societal resilience. Some key points of this case are relevant to the allocation of roles and responsibilities across national/regional/local levels of government, outlined below.
3.2. The QRA was set up as part of a response to a major state-wide emergency to fill a gap in governance capability, but over time its remit expanded to what can be described as a ‘resilience broker’ across different levels of governance. Its original primary mandate was to distribute funds made available by national and state government to support local recovery. The intention was to provide coordination and efficiency that could not be achieved through local councils managing their individual programmes alone. The QRA was established as a temporary organisation, but legislation has since been reviewed to make it a permanent entity. This is coupled with the expansion of the QRA’s remit for resilience-brokering. Some organisational characteristics to support this capability include:
3.3. The QRA’s strength is that it connects the different levels: it provides assistance and distributes money to the local level and it carries the responsibility to monitor spending (reporting on this to national government). The QRA offers a form of centralised expertise, but the processes in place also allow for the local communities themselves to develop responses to flooding in their area. Collection of data over time has also improved information available for investment decisions. The development of a database of damage and repair information offers new insight and transparency with respect to network performance across hundreds of miles. The UK has different features of funding and coordination arrangements (e.g. Flood Re and local resilience forums), but it could use QRA’s experience of adaptive change over the past decade as a point of comparison.
4.1. Technology development will undoubtedly aid CNI management, but human factors in governance and operations must receive similar attention. Compelling commercialisation narratives that often support proposals for technology development can downplay/ignore the wider change implications of adopting adopt new technology. Eisenberg et al. issue a stark warning in their review of resilience analytics: “Overemphasis on the benefits of data-driven methods can widen the gap between people making decisions during disruptive events (Users) and researchers developing resilience analytics to support them (Modelers)”. They conclude that development of resilience analytics needs “research on understanding the context-dependent implications of improvisation” i.e. understanding how people make decisions and why.
4.2. Current PhD research within my group involves analysing the structure of organisations in the energy sector with respect to information flow through the organisations. A motivation for this research was to understand the management of multi-hazard risk within these organisations and what information is being used to support those decisions. To address this, a new way to visualise (map) flows of information within the organisation was developed. The maps capture various data sources, key software, and data flows from asset monitoring and operations through to strategic long-term investment. These maps are providing evidence as to why technological innovation does not provide the panacea solution to improving resilience. Any new technology must be considered within the existing structure of the organisation, that is, how the organisation makes decisions, how systems fit together, who has access to information, and who has mandated responsibility. Otherwise, it can fail to deliver results if the costs for wider change and upskilling are not accounted for.
I hold a PhD and an MPhil in Engineering for Sustainable Development (ESD) from the University of Cambridge. This built on my first degree in civil engineering and a Master's in Engineering Management from Canterbury University, New Zealand. After several years in industry in Australia and New Zealand (including involvement in post-earthquake reconstruction of infrastructure networks in Christchurch), I returned to academia to pursue my interest in the role of engineers in society, particularly with respect to the governance and management of disaster risk and the lack of clear decision-making processes (and associated priorities) for doing this well. From 2017–2020 I directed a master’s programme at Cambridge for senior professionals in the construction industry. I now lead teaching in resilience on the MPhil ESD programme and continue to work with industry on sustainability and resilience related projects. I supervise research on themes of climate adaptation, critical infrastructure resilience management and socio-technical issues associated with infrastructure services provision.
6 January 2022
 Discussed in MacAskill K and Guthrie P (2014). Multiple Interpretations of Resilience in Disaster Risk Management. Procedia Economics and Finance. 18: 667–674, DOI:10.1016/S2212-5671(14)00989-7.
 See Linkov I and Trump B (2019). The Science and Practice of Resilience. Springer Nature: Switzerland.
 See Rasmussen J (1997). Risk management in a dynamic society: a modelling problem. Safety Science. 27(2/3): 183–213. See also Woods D (2003). Creating foresight: how resilience engineering can transform NASA’s approach to risky decision making. Testimony on The Future of NASA for the Committee on Commerce, Science and Transportation. Available at: https://www.globalsecurity.org/space/library/congress/2003_h/031029-woods.pdf. [Accessed 06.01.2022].
 A contamination example is described in Leveson N, Daouk M, Dulac N and Marais K (2003). Applying STAMP in accident analysis. Workshop on the investigation and reporting of accidents. Sept. Available at: https://shemesh.larc.nasa.gov/iria03/p13-leveson.pdf. [Accessed 06.01.2022].
 See Shuttleworth A and MacAskill K (2021). Net zero adaptation—a review of built environment sustainability assessment tools. Environmental Research: Infrastructure and Sustainability. 1(2). DOI: 10.1088/2634-4505/ac1c5e.
 Proportion of scheme assessment related to mitigation: 33.8%, 25.0%, 29.5%, related to adaptation (respectively): 17.0%, 3.0%, 4.3%. Ibid.
 Ikonomova M (2021). Retrofitting resilience in healthcare facilities against extreme weather events (unpublished master’s thesis). University of Cambridge: United Kingdom.
 See https://www.raeng.org.uk/global/international-partnerships/engineering-x_/safer-complex-systems/case-studies. This project was completed with M Barendecht (University of Cambridge) and C Tilley (King’s College London).
 p.1883 in Eisenberg D, Seager T and Alderson D (2019). Rethinking resilience analytics. Risk analysis. 39(9): 1870 – 1884. DOI: 10.1111/risa.13328.
 p.1881, ibid.
 This research is led by S O’Brien (not yet published).